Optical disc storage technology was pioneered in the 1960s by inventors who devised numerous ways to store analog and digital signals on reflective discs. Today, optical discs are widely used for computer, audio, and video applications. They provide high storage capacity; for example, with the development of blue lasers, a multilayer Blu-ray disc can hold 50 Gbytes or more. Optical media are ideal for mass distribution of data; the manufacturing cost of optical media is a fraction of the cost per byte of magnetic-disk or solid-state media. The life expectancy of optical discs is much longer than that of magnetic media.
With proper storage, a Compact Disc should last for 100 years or more. Optical storage of digital audio data is also universally employed for motion picture film soundtracks on optical film. The design of optical disc systems such as CD, DVD, and Blu-ray, as well as fiber-optic systems, rests on the fundamental principles of optics.
Optical Phenomena
Light is an electromagnetic vibration that can be characterized by wavelength, frequency, propagation velocity, propagation direction, vibration direction, and intensity. The optical spectrum, within the context of the electromagnetic spectrum, is shown in FIG. 1. Light ranges from 7.5 × 10^10 to 6.0 × 10^16 Hz. Wavelength is the distance between identical points on a waveform; the wavelengths of visible light extend from about 400 to 800 nm. Wavelength equals velocity of propagation divided by frequency. Likewise, velocity of propagation is the product of wavelength and frequency. The velocity of light in a vacuum is 3 × 10^8 m/s; light travels slower in other materials. Frequency measures the number of vibrations per second; the frequency of an electromagnetic wave is constant, and does not change in matter, but its velocity and wavelength are reduced. Light at different wavelengths travels at different velocities in the same medium. Light can exist at a single wavelength (monochromatic light) or a mixture of various wavelengths (for example, natural light).
The intensity of a light wave is the amount of energy that flows per second across a unit area perpendicular to the direction of propagation. It is proportional to the square of the amplitude and to the square of the frequency.
FIG. 1 The electromagnetic spectrum, showing the wavelengths used in optical
storage and fiber-optic systems.
FIG. 2 When white light passes through a prism, refraction occurs at both
surfaces. Refraction disperses the beam into its component wavelengths. Similarly,
water droplets in the air act as prisms, creating rainbows.
Refraction occurs when light passes into a medium with a different index of refraction; light changes speed, which causes a deflection in its path. For example, when light in air strikes a glass prism, the velocity of propagation is decreased, the wavelength in the denser medium is shorter, thus the light travels through the receiving medium at a different angle, as shown in FIG. 2. When light leaves the prism, refraction occurs again. When white light (comprising many wavelengths) strikes a prism, each wavelength changes speed differently, and is refracted at a different angle. Light emerges from the prism dispersed into colors of the visible spectrum. The index of refraction is smallest for red light and increasingly larger for smaller wavelengths.
A medium's index of refraction (n) is the ratio of the light's velocity (c) in a vacuum, to its velocity (v) in a medium; in other words, n = c/v. The index of refraction of light in a vacuum is 1.00; in air is 1.0003 (rounded to 1.0); in water is 1.33; in glass is 1.5. The indexes of refraction of the incident and receiving mediums determine the angle of the refracted beam relative to the incident beam. The angle of incidence is the angle between the incident ray and the normal, an imaginary line perpendicular to the materials' interface; the angle of refraction is the angle between the refracted ray and the normal.
When light passes into a medium with higher index of refraction, it is refracted toward the normal, as shown in FIG. 3A. Conversely, when light passes into a material with a lower index of refraction, it refracts away from the normal, as shown in FIG. 3B. As the angle of incidence increases, the angle of refraction approaches 90°; this is the critical angle, as shown in FIG. 3C.
At any incident angle greater than this critical angle, no refraction takes place, and all light is reflected back into the first medium at an angle identical to that of the incident angle, as shown in FIG. 3D. Specifically, when the angle of incidence is greater than the critical angle, the result is total internal reflection. This phenomenon is essential for light propagation in fiber optics, as described in Section 13.
FIG. 3 Light refracts when it passes from one medium to another with different
indexes of refraction. A. When n1 is less than n2, the refracted ray is bent
toward the normal. B.
When n1 is greater than n2, the refracted ray is bent away from the normal. C. When n1 is greater than n2 and the incident angle equals the critical angle, light does not enter n2. D. When n1 is greater than n2 and the incident angle is greater than the critical angle, all light is reflected back into n1; this is total internal reflection.
More specifically, Snell's law states:
where n1 = index of refraction of incident medium (such as a fiber-optic core) 1 = angle of incidence n2 = index of refraction of receiving medium (such as a fiber-optic cladding) 2 = angle of refraction
The critical angle of incidence is critical
where 2 = 90°:
Light is totally reflected at angles greater than critical and the angle of total reflection equals the angle of incidence.
Diffraction
Light can be modeled to propagate via rays (which exist only in theory) in the direction of propagation. Surfaces perpendicular to the rays are called wavefronts (which exist in actuality). Wavefront normals, similar to rays, indicate the direction of propagation. The advance of light can be viewed as a summed collection of an infinite number of spherical waves emitted by point sources; this is described in Huygens' principle. These spherical waves are in phase, thus creating a wavefront. However, destructive interference occurs between the spherical waves at all other angles.
When a wavefront passes through an aperture that is small relative to the wavelength (an order of a half wavelength), diffraction occurs, and the wavefront emerges as a point source. For example, diffraction occurs when a single slit is placed in a barrier. A diffraction grating, invented by Joseph von Fraunhofer in 1821, contains a series of identical equidistant slits. Because of interference, a wavefront will only leave the grating in directions where light from all the slits is in phase. This happens straight ahead, and at certain other angles. The angle of the first oblique wave is a function of the wavelength of the light, and the spacing of the slits. If the spacing of the slits is reduced (the spatial frequency is increased), the oblique ray will leave at a greater angle to the centerline. Similarly, the smaller a physical object, the larger the angle over which light must be collected to view the object; fine detail can be resolved only if the diffraction wavefront is collected by the lens.
As described in the following sections, this angle is specified as the numerical aperture of the lens. A diffraction pattern shows the maxima and minima intensities corresponding to the phase differences resulting from different path lengths. The central maximum is the zero order maxima. Other light is diffracted into a series of higher-order maxima. The diffraction pattern formed behind a circular aperture is known as the Airy pattern, proposed by British astronomer George Airy in 1835.
Specifically, even if a lens is perfectly free of aberration, a laser beam, for example, cannot be focused to a point with a circular lens of finite aperture. The focused laser spot is actually an area where the intensity of light varies as an Airy pattern function, as shown in FIG. 4. This is a circular diffraction pattern of maximum and minimum light intensities corresponding to phase differences. The central spot is the zero-order maximum, and the surrounding rings are higher-order maxima. About 83% of the total light falls in the central spot, and the brightest intensity in the first ring is only 1/60 that of the central spot.
The size of the Airy pattern is determined by the light wavelength and the numerical aperture of the lens. Laser pickup optics are said to be diffraction-limited as opposed to tolerance-limited. The size of the spot is dictated by diffraction; a higher-quality lens would still result in the same spot size. As a result, for example, an optical microscope cannot image the diffraction-limited pit spiral on an optical disc (it would show dark spots, not the three-dimensional contour of the pits); instead, a scanning electron microscope is needed. Similarly, a laser pickup operating at a short wavelength is needed to read the data pits on an optical disc.
FIG. 4 An Airy pattern is a diffraction pattern for a point source emerging
from an aperture. This figure shows the pattern from a circular aperture.
A simple hands-on experiment can be used to demonstrate diffraction. In particular, we can show that the CD data surface causes diffraction, and we can use the result to calculate a CD's track pitch. Fetch a CD and a ruler, and seat yourself with a light bulb--60 W or so--about a meter behind you. Hold the CD about 1/3 meter from your face (reflective side facing you) and angle it so the light bulb's reflection disappears in the disc center hole.
Now slowly move the disc toward your face, keeping the bulb's reflection centered in the hole. The brightly colored display is a result of the diffraction created by the pit track.
Now move the disc away from you until the violet ring is located on the edge of the CD. If you measure this, you might find the disc to be, for example, 20 cm from your eye.
Now you can calculate the track pitch using the equation:
where d = track pitch
= wavelength of the violet light (450 nm)
= distance between eye and disc (e.g., 20 cm)
r = radius of the violet ring (5.5 cm)
Calculating, in this example you will obtain a track-pitch value of 1.7 µm-only 1/10,000,000 of a meter away from the actual pitch (on most CDs) of 1.6 µm.
Resolution of Optical Systems
As described in previous sections, whenever parallel light passes through a circular aperture, it cannot be focused to a point image, but instead yields a diffraction pattern in which the central maximum has a finite width-an Airy pattern. The aperture of an objective lens system is circular.
Consider the images of two equally bright point sources of light. Now, consider each point as an Airy pattern. If the points are close, the diffraction patterns will overlap. When the separation is such that it is just possible to determine that there are two points and not one, the points are said to be resolved, as shown in FIG. 5. This is the case when the center of one diffraction pattern coincides with the first minimum of the other. The distance between the two diffraction pattern maxima equals the radius of the first dark ring. This condition is called the Rayleigh criterion.
Similarly, the smaller the object that is viewed, the greater the angle over which light must be collected. In other words, the diffracted wavefront must be collected. The resolving power of the lens is determined by its numerical aperture (NA). The numerical aperture of a lens is the diameter of the lens in relation to its focal length and describes the angle over which it collects light. NA is calculated as follows:
NA = n sin ?max
where max is the angle of incidence of a light ray focused through the margin of a lens and n = index of refraction of the medium.
A lens is a spatial lowpass filter and the cutoff of the spatial frequency is determined by the numerical aperture; the modulation transfer function (MTF) describes the response of the lens. This defines the minimum size of a resolved object for a given wavelength. Because the finite NA results in a cutoff of spatial frequency, the Airy pattern is analogous to the impulse response of a lowpass filter. An impulse (Dirac delta function) input to an electrical circuit yields its impulse response. In an optical system, the spatial equivalent of the impulse is approximated by a pinhole aperture. The impulse response can be viewed as the cross section of the intensity of the spot-an Airy pattern.
The spot size (d) is often defined as the half-intensity diameter of the Airy pattern such that d = 0.61 /NA, where is the wavelength of the laser light in a vacuum. For high data density on an optical disc, d must be small. For example, in the CD format, is fixed at 780 nm; thus NA must be as large as possible. However, as NA is increased, tolerances become severe: the depth-of-focus tolerance is proportional to NA-2, skew (disc tilt)-tolerance is proportional to NA-3, and disc-thickness tolerance is proportional to NA-4. Clearly, for system stability, the NA should be as low as possible. Balancing these factors, for example, the CD designers selected NA = 0.45 (NA is a dimensionless quantity). Thus, CD spot size is approximately 1.0 µm. The DVD format uses of 635 or 650 nm, and NA of 0.60, whereas the Blu-ray format uses of 405 nm and NA of 0.85. When NA exceeds about 0.5, the spot intensity pattern begins to deviate from the Airy pattern, but in relatively minor ways.
FIG. 5 The Rayleigh criterion determines the resolution of an optical
system. A. Overlapping diffraction patterns cannot be resolved. B. Partially
overlapping patterns are resolved (Rayleigh criterion). C. Non-overlapping
patterns are resolved.
Similarly, and NA determine other specifications such as track pitch, cutoff frequency, and track velocity; ultimately, from these parameters, disc storage capacity can be deduced. For example, the spatial cutoff frequency in a CD system can be determined:
Thus formations with a higher spatial frequency (for example, lines smaller than 1.15 lines per micrometer) cannot be resolved. Optical systems must be designed to operate within this constraint. For example, in the CD system, the shortest pit/land length is 0.833 µm. A series of short pit/lands would thus yield a spatial frequency of 1/(2)(0.833 × 10 6) = 0.600 × 106, or about half the cutoff frequency and hence would be easily readable.
Furthermore, given a CD with a track velocity of 1.2 m/s, the temporal cutoff frequency is: In terms of an optical channel, the amplitude of the modulated signal is maximum at 0 Hz and linearly decreases to 0 at 1.38 MHz.
Polarization
Electromagnetic waves such as light waves (and radio waves) consist of electric and magnetic fields that are perpendicular to each other, and both oscillating transversely to the direction of propagation. (The reader will recall that sound propagates longitudinally.) Either field can be used to characterize light, but the electric field is generally used. The transverse electric field can oscillate with vertical, horizontal, diagonal, or other more complex geometry, but is always perpendicular to the direction of travel. Light from a light bulb is said to be unpolarized because an infinite number of electric fields exist and are randomly perpendicular to the direction of travel. Any one of these electric (E) fields, at any angle, can be considered as a vector represented by two orthogonal components Ex and Ey. Examination shows that the component waves are phase-incoherent.
However, when only one electric field is allowed to oscillate, in a single direction that is perpendicular to the direction of travel, the light is said to be linearly polarized or plane polarized. A polarizer accomplishes this, as shown in FIG. 6. The electric field lies in one plane, and does not change direction; moreover, one plane contains the electric vector and the direction of propagation. When the orthogonal components representing the E field are examined, we observe that each of them represent linearly polarized light, and are in phase with each other.
Combined, they produce a single linearly polarized wave. In addition, the linearly polarized Ex and Ey components can be combined with a relative phase shift. In particular, when the phase difference is 90° (phase quadrature), the light is circularly polarized. The magnitude of the resulting electric vector is constant, but instead of remaining in one plane, it varies from point to point and from time to time.
Specifically, it rotates through a helix once per wavelength; viewed from the direction of propagation, it traces a circle.
At other component phase shifts, the light is elliptically polarized; the electric vector varies as it rotates, and traces out an ellipse. In any case, the wave can be resolved into its linearly polarized components.
FIG. 6 A polarizer is used to create linearly polarized light. Only light
waves with an electric (E) field vector parallel to the transmission axis can
pass through. The E vector can be represented by two components Ex and Ey.
Isotropic materials have one velocity of transmission independent of the plane of propagation. A material is anisotropic when the light entering it does not have the same velocity in all planes; in other words, the index of refraction depends on the direction of travel through it.
Anisotropic materials such as calcite can be used to create polarized light. When unpolarized light enters an anisotropic medium, light rays split into two part rays. One ray passes through the object as in an isotropic medium, diffracting normally; the second ray is refracted more strongly and is thus displaced from the first as it emerges.
The first ray is called the ordinary ray, and the second is called the extraordinary ray. The two rays are linearly polarized in mutually perpendicular planes. This is known as double refraction or birefringence. The direction along which no birefringence occurs, where the material behaves exactly like an isotropic medium, is called the optic axis.
The plastic used in optical-disc substrates can exhibit birefringence after it is subjected to the stress of melting and injection molding during disc manufacture. This birefringence is an unwelcome effect of anisotropy. A wavefront (perhaps emitted from a laser pickup) traveling through the substrate is distorted because the two orthogonal components travel at different velocities, creating a birefringent image. The velocity difference depends on the direction of the light ray passing through the birefringence material. In practice, birefringence is minimized through careful manufacturing techniques. With shorter laser wavelength and higher NA, birefringence increases and its effects become more severe. On the other hand, thinner substrates (such as in DVD) reduce birefringence and optical aberrations.
A rotation in the plane of polarization, resulting from different velocities in different planes, can be usefully employed. A retardation plate is a slice cut from a crystal in such a manner that the slice contains the optic axis. A beam of unpolarized light normally incident on the plate will create an ordinary and extraordinary beam. The phase difference between these beams is proportional to the distance traveled within the plate. When the beams emerge, they are not separated, but they are out of phase.
If the thickness of the plate is such that the phase difference at emergence between the superimposed ordinary and extraordinary beam is /4, the plate is called a quarter-wave plate (QWP). Similarly, if the phase difference is /2, the plate is a half-wave plate.
By passing linearly polarized light through a QWP, it can be converted to circularly or elliptically polarized light, depending on the angle between the incident vibration plane and the optic axis. If the angle between the plane of linear polarization and the optic axis is exactly 45°, light is transformed from linear to circular polarization (or vice versa), as shown in FIG. 7. For other angles, the transformation is from linear to elliptical (or vice versa).
This rotation in the plane of polarization can be used to distinguish between light beams in an optical disc pickup.
For example, a linearly polarized light beam might pass through the QWP to become circularly polarized, strike an optical disc, and return through the QWP again becoming linearly polarized. Because the resulting plane of polarization is perpendicular to the incident linearly polarized beam, it can be separated from the incident light by a polarizing prism, acting transparently to the incident light, but as a mirror to the reflected light. Many optical pickups use this technique, as described in Section 7. A half wave plate operates similarly, but the plane rotates at twice the angle of the plane of polarization of the incident beam.
FIG. 7 Linearly polarized light passing through a quarter-wave plate is
converted to circularly polarized light.
The angle between the plane of linear polarization and the optic axis must be 45°.
Design of Optical Media
Most optical storage systems store data across the surface of a flat disc. This allows random access of data, as well as ease of manufacturing replication. In most designs, because data is written and read via optical means, there is no physical contact between the pickup and the media.
This ensures long pickup and media life and minimizes damage from head crashes or other catastrophes. In addition, because there is no need for physical contact between the pickup and the data surface, data can be embedded within a transparent protective layer to minimize the effect of surface contamination and damage on data readout. Also, multiple data layers can be placed within one substrate. However, stored data must undergo both modulation and error-correction encoding to maximize data density and guard against errors.
Data can be stored along either a spiral track or concentric tracks. Most optical disc pickups shine a laser on the media, and the reflected light is detected by a sensor and decoded to recover the carried data. To accomplish this, the media must present two states so the change between them varies the reflected light, and thus the data can be recognized in much the same way that the black characters on this page stand in contrast to the white paper. Data can be represented as a phase change, polarization change, or change in the intensity of reflected light. For example, pits in a reflective surface produce diffraction in the reflected light beam, decreasing its intensity. The resulting variation in intensity, from high intensity in reflecting areas, to low intensity in pit areas, can be converted to a varying electrical signal for data recovery.
To expedite data writing and recovery, a laser beam is used. Laser light, unlike incoherent light, yields a high signal-to-noise (S/N) ratio from the photodetector. The short wavelength permits a high information density to be achieved; for example, a data pit length might be 0.2 µm.
As noted, track pitch, minimum spot size, and other dimensions are defined by the wavelength of the reading laser. Discs are thus diffraction-limited. Shorter wavelength laser light would yield greater storage density. A laser light source is also required to provide a sufficient S/N ratio for a high bit rate, given an illumination of an area on the order of 1 µm2, for a period of 1 ms or less. Either analog or digital signals can be encoded and stored on a disc, within the bandwidth constraints of the media.
Any optical media must be supported by a sophisticated servo system to provide positioning, tracking, and focusing of the pickup, as well as accurate disc rotation. Focus tolerance, for example, can be on the order of 1 µm, and should be maintained in spite of mechanical shock and vibration. The pickup must generate a set of correction signals derived from the signal from the optical media itself and use a set of actuators to maintain proper conditions.
Radial tracking correction signals can be derived using methods such as the twin spot, which uses a diffraction grating to create additional scanning spots, or wobble, in which the scanning spot is given a sinusoidal movement.
Similarly, focus correction signals can be generated through methods such as the Foucault knife edge, using unequal distribution of illumination in a split beam; astigmatism, using a cylindrical lens to create spot asymmetry; or critical angle, using angle of incidence and a split beam. Laser pickup design is discussed in more detail in Section 7.
The shelf life of optical discs is longer than that of magnetic media, optical discs are less susceptible to damage from heat, and they are impervious to magnetic fields. The raw (uncorrected) error rate of an optical disc can be 10^-6, perhaps 10 to 30% of the disc capacity may be needed for error-correction coding to bring the corrected rate to 10^-13. Any optical recording material must exhibit long-term stability, high absorptivity at the recording wavelength, low writing energy, high S/N ratio, good forming characteristics, low thermal conductivity, and low manufacturing cost. It is advantageous to design optical storage with specific applications in mind. Specifically, three separate systems, read-only, write-once, and erasable media have been developed for various applications.
Non-erasable Optical Media
In non-erasable optical disc systems, a laser light shines on the data surface, and the reflected light is detected by a photodiode and decoded to recover the carried data. To accomplish this, the surface presents two states, for example, to vary the intensity of the reflected light. In this way, data can be recognized by the pickup. For example, a reflective disc might have pits embossed into its surface to vary reflected intensity. The actual technology used depends on whether the media is read-only or write-once.
Several mechanisms are available to achieve both of these results, but questions of durability, density, and feasibility of mass production ultimately define the most appropriate method.
Read-Only Optical Storage
CD-Audio, CD-ROM, DVD-Video, DVD-Audio, DVD ROM, and Blu-ray BD-ROM discs are examples of read only optical media. Whether a disc holds audio or software data, the data is permanently formed on the media during manufacture; it is a playback-only format. A plastic disc is impressed with a spiral track of pits set to a depth calculated to decrease the intensity of the laser light of the reading pickup. To provide reflectivity, the data surface is metallized. The reflective surface of the disc and the data pits are embedded between the transparent plastic substrate and a protective layer. The effects of scratches and dust particles on the reading surface are minimized because they are separated from the data surface and thus made out of focus with respect to the laser beam focused on the inner reflective data surface.
In a read-only optical media, the pit surface and laser readout forms a sophisticated optical system. The numerical aperture of the lens, wavelength of the laser light, thickness and refractive index of the disc substrate, and size and height of the pits all interact. As noted, when a laser beam is focused on the data surface, an Airy pattern results, as shown in FIG. 8. The spot diameter can be specified at half-intensity (for example, 1.0 µm) or at the first dark ring at d = 1.22 /NA (for example, 2.1 µm).
Allowable crosstalk between tracks determines track pitch (for example, 1.6 µm). Crosstalk must be acceptable even in the worst case of a slightly defocused beam and slight tracking error. In any case, the beam is focused so that approximately half its center area falls on a pit, and half falls on the surrounding reflective land.
Many theories are employed to model the readout of optical discs. Perhaps the most direct approach models the disc as a diffraction grating. An optical storage surface (such as a read-only CD, DVD, or Blu-ray) that uses a phase structure is a reflective phase grating, and acts similarly to a diffraction grating. Specifically, the pits cause diffraction. The smaller the pit in relation to the light wavelength, the greater is the angle at which light leaves.
The area of light striking a pit is about equal to that striking the surrounding land. Using a simple model, light diffracted by the grating consists of a single zero-order and multiple first-order beams, as shown in FIG. 9. These beams partly overlap each other; the resulting destructive interference between the zero- and first-order beams in the light ray returning to the lens yields cancellation due to interference. Effectively, a pit thus reduces the intensity of light returning to the lens. The image of the interference pattern from the multiple reflection orders is sometimes called the "baseball" pattern. For good contrast, the power in the two beams with different phases should be equal. Pit depth does not require great accuracy. Equality of pit/land areas is more important. A plane-wave model (with a large magnification system) predicts that pit height should be one-quarter of the apparent wavelength of the light, so that light striking a pit is out of phase by a half wavelength with that striking the land. A more complex spherical-wave model that accounts for effects of the converging focused beam (low magnification), devised by Charles Mecca, Yajun Li, and Emil Wolf, shows that the optimum pit depth should be one-half wavelength.
FIG. 8 In write-once optical disc systems, the reading laser forms an
Airy pattern on the data surface. The half intensity spot diameter covers the
pit track.
FIG. 9 In write-once optical disc systems, the data surface forms a phase
structure that acts like a reflective phase grating. Diffraction causes cancellation
through interference in the reflected beams.
The CD, DVD, and Blu-ray systems are diffraction limited because they operate at dimensions that are as small as permitted by the wave nature of light at their respective wavelengths. For example, with a CD spot diameter of 1.0 µm, details this size would give diffracted rays that only just fall within the lens aperture. Finer details yield greater convergence thus rays diffracted by the pits must fall outside the lens aperture. This is in fact the case with the pits on the data surface because they are narrower than the diameter of the spot. Rays diffracted by the pits must consequently fall outside the lens aperture, decreasing intensity reflected back into the lens, promoting a robust playback signal.
This type of diffraction media, using physical pits to store data, is attractive because it can be economically mass-produced. For example, on 12-cm-diameter discs, a CD can hold 680 Mbytes of formatted data, a DVD can hold 4.7 Gbytes on a data layer (multiple layers are possible), and a dual-layer Blu-ray disc can hold 50 Gbytes.
The media is open-ended with respect to the type of data to be encoded. For example, the CD-Audio standard was joined by CD-ROM, Video-CD, SACD, and other formats.
Manufacturers have developed higher-density DVD and Blu-ray discs with many times the storage capacity of a CD; this is accomplished with a shorter wavelength laser, higher NA, and thinner substrate, which permits smaller pit and track dimensions. The CD format is discussed in Section 7, the DVD format in Section 8, and the Blu-ray format in Section 9.
Write-Once Optical Recording Recordability is essential for many applications; the simplest recordable optical systems are write-once (WO).
The user records data permanently, until the disc capacity is filled. When only a few discs are needed for distribution, a write-once system is cost-efficient and write-once discs are widely used throughout the computer, and audio and video industries. A write-once optical disc can be implemented in a variety of ways, as summarized in Fig. 10. Dye-polymer recording uses a recording layer containing a heat-absorptive organic dye; the dye is absorptive at the wavelength of a writing laser. When the layer is heated, it melts and forms a depression.
Simultaneously, a reflective layer is deformed. Data is read by shining a laser light on the surface; the physical formations decrease the intensity of the reflected beam.
Several types of disc recorders use a dye-polymer method.
The CD-WO format (better known as CD-R) is discussed in Section 7. The DVD-R format is discussed in Section 8. Some Blu-ray BD-R discs use dye-polymer recording; Blu-ray is discussed in Section 9. Instead of the "pit and land" nomenclature used to describe prerecorded optical media, some texts refer to the "marks and spaces" of recordable optical media.
In some systems, an irreversible phase change is used to alter the reflectivity of the media at the point where a writing laser is focused. In this way, a reading laser can differentiate between data. Some systems use a thin metallic recording layer that varies its physical property from crystalline to amorphous when it is heated by a writing laser. The crystalline state is more translucent, thus more light reflects from the metal layer above, yielding high reflectivity. The amorphous state absorbs the reading laser light, yielding low reflectivity. The phase transition can triple the reflectivity of the recording layer at written spots, thus allowing laser reading of the data. The recording layer can use an antimony selenium (Sb-Se) metallic film and the heat-absorbing layer can be a bismuth-tellurium (Bi-Te) metallic film.
FIG. 10 A variety of methods can be used in write-once storage. A. Pit
formation. B. Bubble formation. C. Dye polymer. D. Phase change. E. Texture
change.
Other writing methods include pit formation, bubble formation, and texture change. With pit formation, a mechanism called ablation uses a laser writer of approximately 10 mW to burn holes in a reflective layer; the material is melted or vaporized by the heated spot. Melting is preferred because no residue is created around the pit.
In either case, data can be read by monitoring the change in intensity. Similarly, bubble formation uses a laser to vaporize a recording layer, causing a bubble to form in an adjacent reflective layer. The bubble can be read with a laser, by monitoring reflecting light levels. The texture change method uses a reflective surface with small aberrations with dimensions and spacing designed to diffract a reading laser. When the layer is heated and melted, it forms a smooth face that increases reflectivity at that point.
Erasable Optical Media
Erasable optical disc systems provide versatile data storage. Data can be written, read, erased, and written again. In most cases, the number of erasures is essentially unlimited. Several recordable/erasable optical media technologies have been introduced in a variety of formats, varying broadly in cost. These erasable technologies include magneto-optical, phase-change, and dye-polymer recording. Magneto-optical recording uses a laser beam and magnetic bias field to record and erase data, and a laser beam alone to read data. Phase-change media use a reversible change in the index of reflectivity of materials.
Dye-polymer media use reversible changes in physical formations induced by heating the recording layer.
Magneto-Optical Recording
Magneto-optical (MO) recording (sometimes known as optically assisted magnetic recording) technology combines magnetic recording and laser optics, utilizing the record/erase benefits of magnetic materials with the high density of optical technology. Magneto-optical recording uses vertical magnetic media in which magnetic particles are placed perpendicularly to the surface of a pregrooved disc. Vertical recording provides greater particle density and shorter recorded wavelengths; however, the high recording density is not fully utilized by conventional magnetic heads because their flux fields cannot be narrowed sufficiently. The recorded area is thus larger than necessary. Optical assistance increases the recording density.
With magneto-optics, a magnetic field is used to record data, but the applied magnetic field is much weaker than conventional recording fields. It is not strong enough to orient the magnetic fields of the particles, thus a unique property of magnetic materials is used. As the oxide particles are heated to their Curie temperature, their coercivity (the minimum magnetic field strength required to reverse an induced magnetization) decreases radically. In other words, when heated, a magnetic material loses its resistance to change in its magnetic orientation; its orientation can be affected by a small applied magnetic field. (Similarly, variations in the earth's magnetic field over time can be traced by studying the magnetic fields imprinted in ancient volcanic rocks.) In the case of MO recording, this allows data to be written with a weak field. For example, coercivity falls almost to zero as the temperature rises to 150°C. A laser beam focused through an objective lens heats a spot of magnetic material to its Curie temperature. At that temperature, only the particles in that spot are affected by the magnetic field from the recording coil, as shown in Fig. 11A. When the beam is turned off, or the area moves away from the beam, it cools below the Curie temperature as the absorbed energy is dissipated to the substrate by thermal conduction. The applied magnetic field is withdrawn, and the magnetic orientation of the field is "frozen" and retained as data. In this way, the laser beam creates a recorded spot much smaller than otherwise possible thus increasing recording density. Moreover, at room temperature, the recording layer's high coercivity makes it highly resistant to the effect of stray magnetic fields.
The Kerr effect is used to read data; it describes the slight rotation of the plane of polarization of polarized light as it reflects from a magnetized material. (The Faraday effect describes the same phenomenon as light passes through a material.) The rotation of the plane of polarization of light reflected from the reverse-oriented regions differs slightly (by perhaps ± 0.5°) from that reflected from unreversed regions. To read the disc, a laser is focused on the data surface, and the angle of rotation of reflected light is monitored, as shown in FIG. 11B. An analyzer distinguishes between rotated and unrotated light, and converts that information into a beam with varying light intensity. For example, by passing the polarized light through a polarizing prism, light with a polarization plane parallel to that of the prism will pass through, while relatively rotated light will be blocked, resulting in an intensity difference. Data is then recovered from that modulated signal. The power of the reading laser is much lower than the recording laser (perhaps by a factor of 10) so the recorded magnetic information is not affected. To erase data, a magnetic field is applied to the disc along with the laser heating spot, and new data is written, as shown in FIG. 11C. New data can be rerecorded.
FIG. 11 Techniques used in magneto-optical recording, reading, and erasing.
A. In magneto-optical recording, a high-power laser beam heats the data surface,
and magnetic recording is induced by a bias coil. B. In magneto-optical reading,
data is read by reflecting a low power laser beam from the data surface and
observing the rotation in the plane of polarization of the reflected light.
C.
In magneto-optical erasing, data is erased by heating the data surface and reapplying the bias field.
The magneto-optical recording layer can be placed between a transparent substrate and a protective layer. The laser light can shine through either the substrate or the protective layer, placing surface dust and scratches out of focus with respect to the data surface. A coil can be wrapped about the laser lens structure to produce the magnetic field. In some cases, its perpendicular alignment is assisted by a metal plate located on the opposite side of the disc, or the coil can be placed on the opposite side of the disc. A variety of magnetic materials can be used, selected on the basis of S/N ratio, orientation properties, and long-term stability. In general, amorphous, thin-film magnetic materials are used. Some applications use a material such as terbium ferrite cobalt. At room temperatures, the coercivity of the recording layer can be more than 10,000 oersteds, effectively eliminating the possibility of accidental erasure at room temperature.
Magneto-optics is thus potentially more stable and reliable than other magnetic media. Tests indicate that MO data can be erased and rewritten 10 million times or more, equivalent to conventional magnetic media. Accelerated life measurements suggest MO discs will last at least 10 years. In one test, when exposing discs to 95°C at 95% humidity for 1000 hours, MO discs survived better than conventional hard disks.
To achieve mechanical compatibility from one recorder to another within the high tolerances of magneto-optical media, blank discs can be manufactured with prerecorded and nonerasable addressing. One method, called hardware address sectoring, uses a disc with spiral or concentric grooves, in which address information is physically formed in the groove and detected by light beam reflection. Using this system, a magneto-optical player can automatically track both address and data information contained on an MO disc. Storage capacity is not reduced because the recorded data signal is superimposed on the hardware addressing information.
The MiniDisc format used a small MO disc to provide recordability and portability. To provide 74 minutes of recording time on the 2.5-inch disc, the Adaptive TRansform Acoustic Coding (ATRAC) data-reduction algorithm was used.
Phase-Change Optical Recording
Erasable phase-change systems use technology similar to that used in write-once systems. They use materials that exhibit a reversible crystalline/amorphous phase change when recorded at one temperature and erased at another.
For erasable media, a crystalline (translucent and yielding high reflectivity from the metal layer) to amorphous (absorptive and yielding low reflectivity) phase change is typically used to record data and the reverse to erase.
Information is recorded by heating an area of the crystalline layer to a temperature slightly above its melting point. When the area rapidly cools below the crystallizing temperature and solidifies, it is amorphous, and the change in reflectivity can be detected. Because the crystalline form is more stable, the material will tend to change back to this form.
Thus, when the area is heated to a point above its glass transition temperature but below its melting temperature and cooled gradually, it will return to a crystalline state, erasing the data. The temperature during melting might reach 800°C. The dielectric layers mediate the thermal stress, and the aluminum layer acts as a heat sink. The phase-change recording mechanism is shown in FIG. 12.
A number of materials have been devised for the recording layer. For example, layers comprising gallium antimonide and indium antimonide have been developed.
Some systems use tellurium alloyed with elements such as germanium and indium. Unlike the organic dyes used in write-once discs, phase-change recording materials are not wavelength-specific. Permanent recording can be achieved by simply increasing the power of the writing laser; this burns holes in the recording layer rather than changing its phase. Phase-change media have a long shelf life and are not affected by ambient temperatures and humidity. A large number of erasures can be achieved.
Phase-change technology is used in the CD-RW format as described in Section 7. The DVD-RW and DVD-RAM formats use phase-change recording and are discussed in Section 8. Most recordable Blu-ray BD-R and BD-RE discs use phase-change technology; the Blu-ray format is discussed in Section 9.
FIG. 12 In phase-change recording, the recording media may be placed in
an amorphous or crystalline state, depending on the heating temperature applied.
Dye-Polymer Erasable Optical Recording
In dye-polymer recording, light-absorbing dyes are placed in a bi-layer disc structure. The two layers, an expansion layer and retention layer, are sensitive to different wavelengths. The physical deformation that the polymer layers undergo results in bumps; in this way, data can be written and read. For example, the top expansion layer can be composed of an elastomer containing a dye sensitive to light with an 840-nm wavelength. The inner retention layer can be sensitive to 780-nm wavelength light. During recording, infrared laser light of 840-nm wavelength is absorbed in the top recording layer, causing its temperature to rise; the bottom layer is transparent to this light. The top layer's heat causes it to expand, pushing against the bottom layer to form a bump. When the temperature cools, the bump is retained. This formation can be read by a low-power laser beam through diffraction.
To overwrite new data, light of a 780-nm wavelength is absorbed by the bottom layer, but not the top layer. The retention layer is softened, reducing its modulus of elasticity. The expansion layer pulls itself back to its original condition, restoring the surface to a flat condition, ready for new data.
The system is efficient because the recording surface changes state each time laser light with the correct wavelength and power strikes the surface. One wavelength creates a bump; another wavelength flattens it. In addition, the wavelength that flattens a bump does not affect an already smooth surface. Data can thus be overwritten on a single revolution of the media, with data passing under the smoothing laser just before it passes under the forming laser.
Digital Audio for Theatrical Film
Several theatrical film formats have been developed to provide conventional optical storage of frame images, as well as multichannel digital audio storage. In double systems, external audio playback devices are synchronized to the picture using a timecode stripe added to conventional motion picture film. Single systems use optically encoded digital audio data on the film itself to produce a modulated signal. The latter approach, although technically more difficult, is preferred. Motion pictures generally use a stereo optical soundtrack (often matrixed for four channels), called the stereo variable area (SVA) printed along the frame edge; some digital systems retain this track for compatibility with legacy motion picture projection systems. In addition, in the event of catastrophic damage to the digital soundtrack, the system can automatically and momentarily switch to the analog soundtracks. New digital audio cinema systems have supplanted the older 70-mm six-track analog audio format that used magnetic striping.
Any theatrical film audio system must provide multichannel playback, minimally with left, right, and center channels, two surround channels for ambience, as well as a subwoofer channel. Other channels can encode foreign language dialogue or nonaudio data for timecode, control of theater curtains, and so on. Of the six (or more) audio channels, five must provide wide audio bandwidth, with the subwoofer channel reproducing signals below 100 Hz; all must provide a high dynamic range. Optical tracks must be robust, able to withstand hundreds or thousands of trips through the projector; raw error rates of 10^-3 are typical for a worn film. Reliable error correction is mandated. In addition, the method must support high-speed copying for mass replication of films. Several imaging dye methods have been developed in which binary data is recorded as a series of transparent and opaque dots, using conventional film dyes and layers. The mosaic data pattern can be placed along the film edges or between the sprocket holes, as shown in FIG. 13. During playback, light from a source is focused on the tracks, and read with a sensor array on the opposite side. To encode sufficient data on the film, data reduction methods must be used, as described in Sections 10 and 11. Film using this method can be copied at high speed using conventional methods.
The Dolby Digital system retains analog soundtracks for compatibility, and adds an optical data track between sprocket holes, on the same side as the analog tracks. Six audio channels (left/center/right in front, left/right stereo surround in rear, and subwoofer) are sampled at 48 kHz, encoded through Dolby Digital (AC-3) data reduction, multiplexed together and with a 9.6 kbps (thousand bits per second) data channel, and written to film as 96 data blocks.
Data blocks are formed from a 76 × 76-pixel array matrix; each pixel is 32 µm2. The composite bit rate is 320 kbps, excluding error-correction data. The data is placed in the green layer of the track negative. A CCD scanner reads the optical information, and then it is demultiplexed, decoded, and output to the theater sound system. Audio data is recorded 2.5 seconds prior to picture to allow time for processing, and so the timebase can be adjusted to achieve subjectively correct synchronization of sound to picture for a given theater size. The subwoofer channel covers a range of 3 Hz to 125 Hz. Printing laboratories are able to perform high-speed duplication using special heads. Dolby Digital (AC-3) is discussed in Section 11. The Dolby Digital system was developed by Dolby Laboratories.
The DTS (Digital Theater Systems) format is a double system using external storage. A timecode track is placed between the picture and the standard analog SVA sound track; this code is used to synchronize external CD-ROM drives which hold audio data. The timecode also contains a film serial number and a reel number that corresponds to a matching code on the CD-ROM disc. As with other double systems, time-code is read in advance of the picture so that any edits in the picture (such as missing footage or reel changes) can be anticipated and compensated for by the sound source. The timecode track can be placed in either the green or red layer of the track negative. DTS discs contain five data-compressed audio channels; the subwoofer track and the 20-Hz to 80-Hz information from the stereo surround channels is summed and recorded in that region of both surround tracks. The surround tracks are thus full-range while the surround speakers reproduce only 80 Hz and above. The subwoofer channel covers a range of 20 Hz to 80 Hz. Audio data is output on an SCSI bus to external compression decoders that drive movie house sound systems. Discs are placed in shipping containers that fit within the standard cases used to ship 2000-foot projection pancakes to theaters. Data compression is performed with the apt-X100 algorithm. The DTS system was developed by Digital Theater Systems.
FIG. 13 Detail showing one edge of a 35-mm film stock holding three digital
audio tracks and one stereo analog audio track: Sony SDDS data on the top edge
(and bottom edge), Dolby Digital data between sprocket holes, analog stereo
SVA track, and a DTS timecode track synchronizing to external CD-ROM drives.
In the SDDS (Sony Dynamic Digital Sound) system, eight audio channels are encoded and placed in two data tracks running outside the perforation holes, one stripe on each side of the film. Existing optical analog tracks are retained. Each stripe comprises 64 rows of 24 × 22-µm pixels; the "P" stripe (adjacent to the picture) carries the left and left-center channels as well as center and subwoofer.
The "S" stripe (adjacent to the optical soundtrack) carries data for the right and right-center channels, as well as center and subwoofer. All audio data is offset by 17.8 frames; the P data leads while the S data is in edit sync with the picture. If a data track is damaged, a concealment mode accesses additional data tracks. The data is placed in the red layer of the track negative. This system adds full range left-center and right-center tracks to the standard 5.1 channel format. Five loudspeakers are placed behind the screen, along with one subwoofer loudspeaker, also usually placed behind the screen, and two surround sound loudspeakers; smaller playback configurations are supported as well. A professional version of the ATRAC data reduction algorithm, originally devised for the MiniDisc format, is used to code the audio channels. The subwoofer channel covers a range of 4 Hz to 500 Hz. The ATRAC algorithm is discussed in Section 11. The decoder also provides digital-domain equalization, level trims, and surround delay trims. The SDDS system was developed by Sony Corporation.